Arylation of Halogenated Pyrimidines via a Suzuki Coupling Reaction

Richard T. Wheelhouse, Sharon A. Jennings, Victoria A. Phillips, Dimitrios Pletsas, Peter M. Murphy, Nichola C. Garbett, Jonathan B. Chaires, and Tere...
3 downloads 0 Views 46KB Size
J. Org. Chem. 2001, 66, 7125-7128

7125

Arylation of Halogenated Pyrimidines via a Suzuki Coupling Reaction Jennifer M. Schomaker and Thomas J. Delia* Malcolm H. Filson Laboratories, Department of Chemistry, Central Michigan University, Mt. Pleasant, Michigan 48859 [email protected] Received June 8, 2001

The Suzuki coupling reaction has been used extensively for the synthesis of a wide variety of unsymmetrical biaryl compounds. We have extended this reaction to demonstrate the utility of preparing monophenyl-, diphenyl-, or triphenylpyrimidine depending on the reaction conditions. Further, it has been shown that chloropyrimidine substrates are preferable over iodo-, bromo-, or fluoropyrimidines. Introduction There have been relatively few methods available to produce C-aryl derivatives of pyrimidine. Perhaps the most practical of these methods involves the primary synthesis in which the pyrimidine ring is constructed by condensation of two components bearing the desired aryl groups in appropriate locations. An example of this approach has been described by Pinner1 to produce 4,6dimethyl-2-phenylpyrimidine from benzamidine and acetylacetone. Likewise 2,4,6-triphenylpyrimidine can be prepared in analogous fashion.2 Recently a modification of this approach has been described,3 utilizing a three component coupling-isomerization sequence, catalyzed by a palladium complex to produce 2,4,6-tri(hetero)arylsubstituted pyrimidines. The direct arylation of pyrimidines is a simpler approach that does not involve the preparation of productspecific intermediates. However, examples of this chemistry are quite limited. The reaction of a phenyl Grignard reagent combining with a suitable pyrimidine is one of the earlier methods described. Thus, phenylmagnesium bromide is reported to react with unsubstituted pyrimidine 1 to afford 4-phenylpyrimidine 24 (Scheme 1), and with 4-chloro-5-ethyl-2,6-diphenylpyrimidine 3 to give the corresponding triphenyl derivative 4.5 In related chemistry, a Negishi coupling of 2,4-pyrimidinyl triflate with p-anisylzinc bromide affords 2,4-dianisylpyrimidine.6 Directed ortho metalation of many nitrogen heterocycles, including pyrimidines, is an alternate method for the introduction of aryl substituents.7 Given our continuing interest in investigating the substitution reactions of 2,4,6-trichloropyrimidine,8 we sought to examine the application of Suzuki coupling reactions with a variety of halogenated pyrimidines. (1) Pinner, E. L. Ber. 1893, 26, 2122. (2) Dodson, R. M.; Seyler, J. K. J. Org. Chem. 1952, 16, 461. (3) Muller, T. J. J.; Braun, R.; Ansorge, M. Org. Lett. 2000, 2, 1967. (4) Brown, D. J. In The Pyrimidines; Interscience Publishers: New York, 1994; p 15. (5) Traube, W. Ber. 1893, 26, 2551. (6) Turck, A.; Ple´, N.; Lepretre-Gaquere, A.; Queguiner, G. Heterocycles 1998, 49, 205. (7) Turck, A.; Ple´, N.; Mongin, F.; Que´guiner, G. Tetrahedron, 2001, 57, 4489.

Scheme 1

The electron deficient nature of the pyrimidine ring makes this system far more reactive in Suzuki couplings compared to the analogous benzenoid halides. For example, although aryl chlorides are generally unreactive toward the oxidative addition of palladium without the use of specialized and expensive ligands, the reaction of 2-chloropyrimidine with 5-indoylboronic acid occurred smoothly using tetrakis(triphenylphosphine)palladium.9 The use of Suzuki coupling conditions has also been employed with 5-halogenated pyrimidines. Treatment of 5-bromo-2,4-dialkoxypyrimidines 5 with thienyl- or selenylboronic acids led to the expected 5-heterosubstituted products 610 (Scheme 1). (8) (a) Delia, T. J.; Stark, D.; Glenn, S. K. J. Heterocyclic Chem. 1995, 32, 1177. (b) Delia, T. J.; Nagarajan, A. J. Heterocyclic Chem. 1998, 35, 269. (c) Delia, T. J.; Meltsner, B. R.; Schomaker, J. M. J. Heterocyclic Chem. 1999, 6, 1259. (d) Schomaker, J. M.; Delia, T. J. J. Heterocyclic Chem. 2000, 37, 1457. (9) Yang, Y.; Martin, A. R. Heterocycles 1992, 34, 1395. (10) (a) Gronowitz, S.; Hoenfeldt, A.-B.; Kristjannson, V.; Musil, T. Chem. Scripta 1986, 26, 305. (b) Peters, D.; Hoenfelt, A-B.; Gronowitz, S. J. Heterocyclic Chem. 1990, 27, 2165.

10.1021/jo010573+ CCC: $20.00 © 2001 American Chemical Society Published on Web 09/15/2001

7126

J. Org. Chem., Vol. 66, No. 21, 2001 Scheme 2

Scheme 3

Results and Discussion Unaware of any detailed study of the selectivity of Suzuki coupling at positions 2 or 4 of the pyrimidine ring, we have examined the reaction of phenylboronic acid 7 with three of the four possible 2,4,6-trihalopyrimidines 8 (a ) Cl; b ) I; c ) F) as well as two 2,4-dihalopyrimidines 9 (b ) I; c ) Br) and 4,6-dibromopyrimidine 10b (Schemes 2 and 3). Our two main objectives in this study were (1) to determine which of the four halogens would provide the best result and (2) to ascertain whether a Suzuki coupling reaction would occur preferentially at either of the nucleophilic sites (positions 2 or 4/6). Halogenated Pyrimidines. Since oxidative addition of palladium into a carbon-halogen bond occurs in the order I > Br . Cl . F, based mainly on the strength of the C-halogen bond, we sought to include representative compounds containing each of the halogen atoms in our studies. The precursors for the iodo-, bromo-, and fluoropyrimidines were the corresponding commercially available chloro compounds. Thus, 2,4,6-trichloropyrimidine 8a was treated with aqueous HI to give the previously unknown 2,4,6-triiodopyrimidine 8b in 90% yield. 2,4,6-Tribromopyrimi-

Schomaker and Delia

dine is not commercially available and the literature procedures for its preparation are not easily reproduced.11 Consequently we opted to examine two isomeric dibromopyrimidines instead. (see below) 2,4,6-Trifluoropyrimidine 8c is readily obtained in 56% yield from 8a using KF and 18-crown-6, as catalyst, in sulfolane. This method is simpler than previously reported syntheses of 8c using AgF12a and KF in a sealed tube.12b 2,4-Diiodopyrimidine 9b was formed in 91% yield upon treatment of the corresponding dichloropyrimidine 9a with aqueous HI. Bromination of 9a with PBr3 under reflux afforded 2,4-dibromopyrimidine 9c in 32% yield. It is worth noting that similar treatment of the trichloropyrimidine 8a failed to provide the corresponding tribromopyrimidine. Bromination of 4,6-dichloropyrimidine 10a under conditions comparable to those above led to the formation of the dibromo compound 10b in 41% yield. Suzuki Coupling Reactions. Reaction of 8b with one equivalent of phenylboronic acid 7, catalytic amounts of palladium acetate and triphenylphosphine, and excess Na2CO3 in a water/ethanol/benzene solvent system at 70 °C led to low yields of coupled products, as determined by the 1H NMR spectrum of the reaction mixture, accompanied by a substantial quantity of impurities. The same reaction conditions in a glyme/water mixture provided low, but improved, yields of a mixture of coupled products. Even though the initial stoichiometry was chosen to obtain the isomeric monosubstituted products, neither of these compounds could be isolated. Instead, only the 4,6-disubstituted product 11 (11%) and triphenylpyrimidine 12 (16%) were obtained. (Scheme 2) There was no indication of the formation of the isomeric 2,4-disubstituted compound. Despite the individual low yields, considering that 7 turned out to be the limiting reagent, the mass conversion was ∼70%. Clearly the iodo groups proved to be so reactive that no selectivity could be achieved. Even the use of a less active catalyst, tetrakis(triphenylphosphine)palladium, gave the same results. A 4-fold excess of 8b gave neither of the monosubstituted products, and when the reaction was carried out at room temperature no reaction occurred. In an effort to limit the sites of reaction and, therefore, improve the selectivity, 2,4-diiodopyrimidine 9b was considered. Suzuki coupling of 9b with 7 gave a mixture of the 4-substituted product 13a in 35% yield and the disubstituted product 14 in 20% yield. Even though the major product was monosubstituted, the selectivity was not synthetically appealing. The reaction of 9c with one equivalent of 7 under Suzuki coupling conditions led to a mixture of 2-bromo4-phenylpyrimidine 13b (58%) and 2,4-diphenylpyrimidine 14 (22%) with no evidence of any appreciable quantity of the 2-phenyl derivative, a possible intermediate in the formation of 14. Again the mass balance is quite good, but the bromo groups are also too reactive to give good selectivity. Similar results are observed when 4,6-dibromopyrimidine 10b was treated under the same conditions affording a mixture of the monosubstituted compound 15 and the disubstituted pyrimidine 16. (11) (a) Golding, D. R. V.; Sinear, A. E. J. Org. Chem. 1947, 12, 293. (b) van der Does, L.; den Hertog, H. J. Tetrahedron Lett. 1972, 2183. (12) (a) Schroeder, H.; Kober, E.; Ulrich, H.; Ratz, R.; Agahigian, H.; Grundmann, C. J. Org. Chem. 1962, 27, 2580. (b) Boudakian, M. M.; Kober, E. H.; Shipkowski, E. R. U. S. Patent 3,280,124, 1966; Chem. Abstr. 1967, 66, 2582.

Arylation of Halogenated Pyrimidines

The trichloropyrimidine 8a was treated with one equivalent of 7 in the glyme/water solvent system using Pd(OAc)2/PPh3 as catalyst. (Scheme 3) A dramatic improvement in selectivity was achieved as 2,4-dichloro-6phenylpyrimidine 17 was isolated in 88% yield, accompanied by small amounts of the di- and trisubstituted products. Interestingly, the selectivity of the coupling led almost exclusively to the 4-substituted isomer. 13C NMR could readily confirm the position of substitution, where the spectrum showed four distinct carbons for the pyrimidine ring indicative of an unsymmetrical substitution pattern. The 2-phenyl derivative would exhibit only three signals in the 13C NMR spectrum for the pyrimidine ring. When 8a was treated with 2 equiv of 7 under similar conditions only the 4,6-disubstituted product 18 was isolated in high yield (88%). This product arises by a subsequent substitution of the 4-phenyl derivative 17 at position 6 rather than at position 2. This result, in addition to our earlier report with substitution of chlorine by phenoxide anions,8b continues a growing list of unusual substitution patterns. Previous work by us8a and others13 describes a more statistical ratio of 4-substitution vs 2-substitution when small anionic or neutral nucleophiles are employed. Treatment of 8a with a slight excess of 3 equiv of 7 under identical conditions afforded the triphenyl derivative 12 in 93% yield. Although aryl fluorides are very unreactive toward oxidative addition of palladium, we were curious as to whether the electron-deficient pyrimidine ring coupled with the strong electron-withdrawing effect of fluorine would allow 2,4,6-trifluoropyrimidine 8c to function as a suitable partner in a Suzuki coupling process. Unfortunately, when 8c was treated in a manner similar to the other halogenated pyrimidines no arylation was observed. The major reaction appeared to be hydrolysis of one or more of the fluorine substituents under the relatively forcing conditions employed. We did not pursue further attempts to obtain the desired Suzuki coupled products since the chloro compound gave a satisfactory synthetic method for arylation of the pyrimidine ring. Conclusions We have synthesized or purchased a series of halogenated pyrimidines and subjected them to the conditions normally employed for Suzuki coupling reactions and obtained C-phenyl pyrimidines. Due to the electrondeficient nature of the pyrimidine ring, the typical iodo and bromo substrates proved to be too reactive to exhibit good selectivity where two or more halogens are present on the pyrimidine. The fluorinated pyrimidine, on the other hand, was completely unreactive toward coupling under comparable conditions. The observations that the halogen atoms (I, Br, Cl) attached at C-2 were replaced last provides a clear order of reactivity, namely position 4 > position 6 > position 2. The chloro pyrimidines proved to be excellent precursors for the synthesis of specific C-aryl pyrimidines in very good yields. Furthermore, since chloro compounds tend to be more available com(13) (a) Winkelmann, W. J. Prakt. Chem. 1927, 115, 292. (b) Boon, W. R. J. Chem. Soc. 1952, 1532. (c) Jacobsen, E. J.; McCall, J. M.; Ayer, D. E.; VanDoornik, F. J.; Palmer, J. R.; Belonga, K. L.; Braughler, J. M.; Hall, E. D.; Houser, D. J.; Krook, M. A.; Runge, T. A. J. Med. Chem. 1990, 27, 1145. (d) Mossini, F.; Maggiali, C.; Morini, G.; Impicciatore, M.; Morini, G.; Moliva, E. Farmaco Ed. Sci. 1984, 39, 189.

J. Org. Chem., Vol. 66, No. 21, 2001 7127

mercially, as well as less costly, this pathway is quite desirable. We are currently examining the scope of the Suzuki coupling reaction using substituted phenylboronic acids.

Experimental Section General. Melting points were measured in open capillary tubes and are uncorrected. Proton and carbon magnetic resonance spectra were recorded at 300 MHz for 1H NMR and 75 MHz for 13C NMR using DMSO-d6 or CDCl3 containing tetramethylsilane as an internal standard. Reactions were monitored by thin-layer chromatography on silica gel containing fluorescent indicator using various combinations of hexane/ ethyl acetate or chloroform/methanol as eluant. The plates used were Eastman Kodak. All mass spectral data were obtained by means of direct insertion probe (DIP) methods. Elemental analyses were performed by Galbraith Laboratories of Knoxville, Tennessee. Chemicals and reagents were purchased from Fisher or Aldrich Chemical Companies and were used without further purification. 2,4,6-Triiodopyrimidine (8b). A 100 mL round-bottom flask was charged with 2,4,6-trichloropyrimidine 8a (5.0 g, 27.3 mmol) and 40 mL of HI as a 57% solution in water. The orangecolored bi-phasic system eventually changed to a thick, light orange slurry. After stirring overnight, the slurry was diluted with water, cooled in an ice/water bath and filtered. The resulting solid was washed well with cold water and dried to yield the crude product (11.3 g, 24.7 mmol, 90% yield) as a cream solid. The crude material was recrystallized from hexane to yield pure 8b as ivory-colored needles, mp 203205 °C (7.9 g, 70% recovery, 63% yield); 1H NMR (DMSO-d6) δ 8.3 (s, 1H); 13C NMR (DMSO-d6) δ 127.2, 130.1, 141.6; MS (DIP) m/z 458 (M+, 100), 331 (M-127, 80). Anal. Calcd for C4HN2I3: C, 10.50; H, 0.22; N, 6.12. Found: C, 10.59; H,